Observer drift refers to a situation in which an observer unintentionally changes how they apply a measurement system over time, often leading to inaccurate data collection.
Example
If an observer is tasked with recording instances of hand-raising in a classroom, they might initially count only when students raise their hand fully. However, over time, they may begin to include partial hand-raises or other gestures, causing a drift from the original criteria and leading to inconsistencies in the recorded data.